Use in Critical Illness

B Ronan O’Driscoll and Rachel Smith

Introduction The Discovery of Oxygen History of Medical Use of Oxygen Reasons to Use Oxygen in Critical Care History of Oxygen Use in Critical Care Evidence of Harm From Excessive Use of Oxygen in Critical Care Survivors of Cardiopulmonary Resuscitation Ventilated General Critical Care Patients Stroke Patients on Critical Care Units Traumatic Brain Injury Patients on Critical Care Units Myocardial Infarction Pilot Studies of Conservative Oxygen Management for Ventilated Critical Care Patients Randomized Trials of in Critical Care Settings Why Have Clinicians Used Too Much Oxygen for Many Decades? Best Practice for Oxygen Use in Critical Illness Overview Immediate Management of Critical Illness in Pre-Hospital Settings Immediate Management of Critical Illness in Hospital Settings Oxygen Use in Critical Care Units Choice of Device The Role of the Respiratory Therapist Special Uses of Oxygen Oxygen as a Driving Gas for Nebulizers Poisoning Hyperbaric Oxygen Medically Managed Pneumothorax Oxygen and Avoidance of Perioperative Wound Infection Future Directions

Oxygen is the most commonly used drug in critical care. However, because it is a gas, most clinicians and most patients do not regard it as a drug. For this reason, the use of medical oxygen over the past century has been driven by custom, practice, and “precautionary prin- ciples” rather than by scientific principles. Oxygen is a life-saving drug for patients with severe hypoxemia, but, as with all other drugs, too much can be harmful. It has been known for many decades that the administration of supplemental oxygen is hazardous for some patients with COPD and other patients who are vulnerable to retention of (ie, ). It has been recognized more recently that excessive oxygen therapy is associated with signifi- cantly increased mortality in critically ill patients, even in the absence of risk factors for hypercapnia. This paper provides a critical overview of past and present oxygen use for critically ill patients and will provide guidance for safer oxygen use in the future. [Respir Care 2019;64(10):1293–1307. © 2019 Daedalus Enterprises]

RESPIRATORY CARE • OCTOBER 2019 VOL 64 NO 10 1293 OXYGEN USE IN CRITICAL ILLNESS

Introduction oxygen was much overused.3 He exposed 3 fallacies that will be discussed further in this review. First, there was a Critical illnesses such as major trauma, pneumonia, sep- belief that exposure to supplemental oxygen at concentra- sis, or heart failure can cause dangerously low blood ox- tions up to 60% was without adverse effects. Second, it ygen levels (ie, hypoxemia), which can cause tissue dam- was thought that patients at risk of developing arterial age and death.1 For this reason, avoidance of hypoxemia is hypoxemia could be protected by administering supple- a central tenet of the emergency medical response to crit- mental oxygen. Third, it was thought that routine admin- ical illness. The first two elements of the “ABC” of emer- istration of supplemental oxygen was useful, harmless, gency care refer to ensuring oxygenation by attention to and clinically indicated on a prophylactic basis. the airway and , and the third element (ie, circu- Since 2002, there have been numerous publications that lation) is also critical to deliver oxygen to the tissues. For have supported Downs’ view that supplemental oxygen these reasons, oxygen therapy became a central compo- has been overused, but clinical practice has been slow to nent of resuscitation and critical illness management dur- evolve. This review will explain the principles underlying ing the 20th century. The prevailing philosophy regarding contemporary oxygen therapy together with an outline of oxygen use during that century was that “more is better.” oxygen use in the management of critically ill patients in There was no awareness in the 20th century that high- the past and guidance for safe oxygen use in the manage- oxygen therapy could be harmful apart from ment of critical illness in the future. The authors will try to special circumstances such as in the care of hypercapnic explain why clinical practice in this area has been so slow patients with exacerbations of COPD, who were known to change and why it has been so difficult to translate since the 1960s to require controlled oxygen therapy, and research findings into routine care. in the care of premature babies, in whom excessive oxygen therapy was recognized since the 1950s as a cause of The Discovery of Oxygen blindness.1,2 There was no reliable way of knowing if a patient was adequately oxygenated until they arrived at the It was thought in the 18th century that the process of hospital and had invasive blood gas measurements. Out- gave off a chemical called phlogiston; it is side of critical care units, blood gas tests could be under- now known that combustion (and animal ) ac- taken only intermittently, and repeated arterial sampling is tually involves the consumption of oxygen. An English traumatic for patients. In these circumstances, it is under- scientist, theologian, and polymath named Joseph Priest- standable that 20th-century clinicians managing patients ley (the inventor of soda water and thus of all fizzy drinks) with critical illness administered oxygen lavishly to avoid isolated a colorless gas by heating mercuric oxide. This hypoxemia, and they had no concerns regarding possible gas allowed candles to burn brighter and for longer, and it harm from hyperoxemia outside of the special groups of prolonged the life of mice in sealed containers. Priestley patients mentioned above. thought he had discovered something he called dephlogis- However, with the advent of ubiquitous cheap and re- ticated air. A similar discovery was made by a Swedish liable pulse oximeters in the early 21st century, alongside chemist named Carl Scheele at about the same time, but increasing knowledge regarding the dangers of hyperox- Priestley got most of the credit because he published first. emia and the lack of evidence of any benefit from aggres- It was the French chemist, Lavoisier, who recognized that sive oxygen therapy, the use of oxygen in critical care in Priestley and Scheele had actually discovered a new gas, the 21st century should be very different from what hap- which he called oxygen. It is fascinating to observe that pened in the previous century. In 2002, in his prestigious Priestley recognized the potential medical uses of oxygen, Donald F Egan Scientific lecture to the 48th International as well as the potential risks of accelerating oxidative pro- Respiratory Congress of the American Association for Re- cesses when he watched candles flare up in this new gas. spiratory Care, John Downs concluded that supplemental He issued the following prophetic words in 1776:

From the greater strength and vivacity of the flame of Dr O’Driscoll and Ms Smith are affiliated with the Department of Re- a candle, in this pure air, it may be conjectured, that it spiratory Medicine, Salford Royal University Hospital, Salford, United might be particularly salutary to the lungs in certain Kingdom. morbid cases, when the common air would not be sufficient to carry off the phlogistic putrid effluvium The authors have disclosed no conflicts of interest. fast enough. But, perhaps, we may also infer from Correspondence: B Ronan O’Driscoll MD FRCP, Department of Respi- these experiments, that though pure dephlogisticated ratory Medicine, Salford Royal University Hospital, Stott Lane, Salford air might be useful as a medicine, it might not be so M6 8HD, United Kingdom. E-mail: [email protected]. proper for us in the usually healthy state of the body; for as a candle burns out much faster in dephlogisiti- DOI: 10.4187/respcare.07044 cated than in common air, so we might, as may be

1294 RESPIRATORY CARE • OCTOBER 2019 VOL 64 NO 10 OXYGEN USE IN CRITICAL ILLNESS

said, live out too fast, and the animal powers be too the main fuel for the of the cells and organs of soon exhausted in this pure kind of air. animals and humans, a major fall in the tissue oxygen level will produce rapid and severe injury to several organs, History of Medical Use of Oxygen especially the brain, which is the most vulnerable organ in hypoxic conditions. is common in critical illness, Attempts to use oxygen for medical benefit began in the and hypoxic brain damage is one of the most feared com- late 18th century within a few years of the discovery of plications of hypoxia. Furthermore, a high proportion of oxygen, and these attempts continued through the 19th patients in critical care units require mechanical ventila- century. As early as 1798, founded the tion, which usually involves the use of supplemental ox- pneumatic institution for inhalation gas therapy in Bristol ygen above the level of room air (ie, 21% oxygen). It is for using oxygen and other gases, including (ie, these reasons that supplemental oxygen is administered to laughing gas) manufactured by the famous scientist, Hum- the majority of patients in critical care units. phrey Davy. For more than 100 years, the most common Patients with medical emergencies who are hypoxemic indication for oxygen use was in the treatment of pulmo- are more likely to die than those with normal blood oxy- nary tuberculosis, which was a common cause of death in gen levels.5 However; it is difficult to know in most cases the 19th century. Unfortunately, supplies of oxygen were whether the increased risk of death is a consequence of the limited at that time, and the intermittent application of low blood oxygen level itself or just a marker of disease oxygen via primitive face masks was probably of little severity. It is self-evident that an episode of pneumonia clinical benefit to cyanotic patients with end-stage pulmo- that is sufficiently severe to cause hypoxemia is more nary tuberculosis, for which there was no effective treat- serious (and more likely to result in death) than a milder ment until the mid-20th century. episode of pneumonia in a normoxemic patient. What is Medical oxygen first came into widespread use in the not known is the effect of oxygen therapy on the likeli- management of soldiers who had inhaled toxic gases dur- hood of survival in these circumstances. The level of hy- ing the . The pioneer of this work was a poxemia that is dangerous to human life is unknown and Scottish physiologist, John Scott Haldane, who also in- probably varies depending on the cause of hypoxemia, the vented the . Like Priestley, Haldane had a pre- speed of onset of hypoxemia, and the underlying condition monition that oxygen treatment would be harmful if over- of the patient. Pending future studies, it is believed that used. He wrote the following in his historic publication patients with serious illness and oxygen saturation Ͻ 90% regarding oxygen use in 1917: “The probable risks of pro- are likely to benefit from supplemental oxygen therapy.1,6,7 longed administration of pure oxygen must be borne in However, as discussed in detail below, there is little evi- mind, and if necessary balanced against the risks of allow- dence of benefit from the administration of oxygen to ing the oxygen want to continue.” patients with S above this level, and there is evidence of pO2 The prophetic words of Priestley and Haldane went un- harm if the patient is rendered hyperoxemic by the use of heeded over the course of the 20th century, and oxygen supplemental oxygen. came to be used not just for patients with potential hypox- Oxygen was used in the past as a routine supplemental emia but for virtually every type of medical emergency to measure with little thought regarding the purpose of this the extent that the presence of an oxygen mask became an treatment. However, as discussed below, it is now recog- essential prop in every movie scene that involved acute nized that giving too much oxygen may be as dangerous as illness. An audit of oxygen use in United Kingdom am- giving too little oxygen, although the consequences of bulances in 2008 found that 34% of ambulance patients hyperoxemia are not as immediate nor as obvious to cli- were given supplemental oxygen therapy, even though only nicians as the consequences of sudden hypoxemia. 17% had oxygen saturation Ͻ 94% and only 7% were significantly hypoxemic with saturation Ͻ 90%.4 History of Oxygen Use in Critical Care

Reasons to Use Oxygen in Critical Care Intensive care units and blood gas analysis were first developed in Copenhagen in the 1950s in response to a Many serious illnesses result in hypoxemia due to im- polio epidemic, which caused hundreds of cases of respi- paired in the lungs, which in turn may be ratory failure. The specialty of intensive care was further caused by a variety of factors, including direct pulmonary developed in the United States and in Europe in the 1960s injury (eg, pneumonia or major trauma) and indirect fac- and 1970s.8 The first ventilators were “iron lung” curaisse tors such as disordered hemodynamics in sepsis, shock, devices that did not necessarily involve oxygen therapy. cardiac arrest, or pulmonary embolism. Hypoxia may also These were soon succeeded by mechanical ventilators, be the result of reduced ventilation as a consequence of which required entrained gases. Invasive ventilation al- opioid analgesics or painful respiration. Because oxygen is most always involved the use of oxygen above the envi-

RESPIRATORY CARE • OCTOBER 2019 VOL 64 NO 10 1295 OXYGEN USE IN CRITICAL ILLNESS ronmental level of 21%. Intensive care units also devel- P among 152,680 ventilated critical care subjects in aO2 oped the capacity to artificially support other organ systems Australia and New Zealand was 20.3 kPa (152 mm Hg) including renal replacement, cardiovascular support and over the period of 2000–2009, although the normal range enteral and parenteral nutrition. is just 75–100 mm Hg (10.5–13.5 kPa).10,11 An audit of The early decades of critical care medicine involved 17,292 blood gas samples from the ICU at Salford Royal increasingly aggressive measures to “normalize” the pa- Hospital in the United Kingdom in 2015 showed a mean tient’s and biochemistry. Sadly, most of these oxygen saturation of 95.7%.12 The percentage of samples measures, such as aggressive blood transfusion, are now with saturation Ͼ 98% fell from 57% in 2005 to 43% known to have been harmful. Recent research has shown in 2010, and it fell further to 29% in the repeat audit that this is also true for aggressive or unnecessary oxygen in 2015. In 2005, the mean P of Salford critical care aO2 therapy. patients was 15.1 kPa (113 mm Hg). This was reduced to Most critical care patients and virtually all ventilated 14.9 kPa in 2010 and to 13.5 kPa in 2015. These findings critical care patients are administered supplemental oxy- suggest that oxygen use in critical care has become more gen at a higher concentration than ambient air, which con- conservative in recent years, which is in line with a grow- tains 21% oxygen. Supplemental oxygen therapy can range ing literature suggesting possible harm from very high from 24% to 100% oxygen, but the safest oxygen level for blood oxygen tension as discussed below. Recent guidance critical care patients is not known, and the optimal oxy- for ventilated patients on critical care units recommends genation goal should probably be adjusted to the patient’s aiming for normoxemia or consideration of permissive hy- specific circumstances. For example, it is generally agreed poxemia. Evidence for the latter strategy is limited and that patients with acute lung injury should have lower mainly based on observational data.13 oxygenation targets than other critical care patients.9 Un- fortunately, we do not yet know which levels of oxygen- Evidence of Harm From Excessive Use of Oxygen in ation are best for most medical conditions. Critical Care Oxygen levels in critical care units are monitored with The high prevalence of elevated blood oxygen tensions the use of continuous pulse oximetry, which displays a in critical care units in several countries (discussed above) percentage (S ). Pulse oximeter values of 94–98% are pO2 is of concern because a number of retrospective observa- generally taken to be normal, and values Ͻ 90% have been tional studies and one prospective randomized study have considered as cause for concern.1 Blood gas measurements, identified increased mortality in association with hyperox- which include arterial oxygen and carbon dioxide tensions emia as well as the expected increase in mortality that is (P and P ), in addition to blood pH, bicarbonate, and aO2 aCO2 associated with hypoxemia. lactate levels, are also measured intermittently in samples There is evidence that high fractions of inspired oxygen of arterial blood drawn from intra-arterial catheters, direct are directly harmful to lung tissue. In the 19th century, radial artery sampling, or arterialized capillary samples.1 exposure to 73% oxygen at atmospheric for 4 d A consensus conference on mechanical ventilation in was reported to cause fatal pneumonitis in rats, and similar 1993 concluded as follows: problems have been described in mice, likely mediated by 14–16 A critical objective of mechanical ventilation is to enhanced reactive oxygen species activity. Griffith achieve and maintain a level of arterial blood oxygen- et al reported in 1986 that breathing 50% oxygen for 45 h ation that is acceptable for the clinical setting, using an led to evidence of pulmonary leakage and inflammation in F that is also acceptable. In most applications of normal human volunteers.17 IO2 ventilatory support, this means a S Ͼ90% (roughly aO2 Oxygen is also known to lead to atelectasis and coro- equivalent to a P Ͼ60 mm Hg or 8 kPa, assuming 18,19 aO2 nary and cerebral vasoconstriction. In addition to the a normal position of the oxyhemoglobin dissociation direct and indirect effects of oxygen therapy itself, some curve), although other end points are appropriate in patients are difficult to ventilate, and it is possible that certain settings. There is no clinical evidence that a aggressive strategies used to “normalize” the oxygen sat- P greater than normal is advantageous.6 aO2 uration (eg, high inflation ) may themselves be harmful to the lungs. A number of retrospective studies and 4 systematic reviews have concluded that hyperox- Despite this guidance from 26 years ago, there was an emia is consistently associated with increased in-hospital ongoing climate until quite recently where blood oxygen mortality in several subsets of critically ill subjects.20-23 levels at the top end of the normal range or in varying degrees of hyperoxemia were the norm on critical care Survivors of Cardiopulmonary Resuscitation units throughout the world. For example, the mean P of aO2 36,307 ventilated critical care subjects in the Netherlands Kilgannon et al24 reported an increased risk of in-hos- in 1999–2006 was 12.4 kPa (93 mm Hg), and the mean pital mortality (adjusted odds ratio for death ϭ 1.8) for

1296 RESPIRATORY CARE • OCTOBER 2019 VOL 64 NO 10 OXYGEN USE IN CRITICAL ILLNESS

6,326 subjects after cardiopulmonary resuscitation who Stroke Patients on Critical Care Units had P Ͼ 300 mm Hg (Ͼ 40 kPa) on their first critical aO2 care unit blood gas measurement. A subsequent paper by Rincon et al29 analyzed 2,894 stroke subjects treated on Kilgannon et al25 showed that this was a linear dose-re- critical care units and reported a crude odds ratio of 1.7 for 26 hospital mortality among subjects with P Ͼ 300 mm Hg sponse relationship. Bellomo et al looked at the lowest aO2 oxygen level (worst A-a gradient) in the first 24 h in the compared with normoxia. After multivariate analysis, the critical care unit for 12,108 post resuscitation subjects and adjusted odds ratio was 1.2 (95% CI 1.04–1.5). In a quasi- they reported a mortality rate of 59% in the randomized trial, increased mortality was also reported in group (P Ͼ 300 mm Hg) compared with a 47% mor- stroke subjects on general wards with mild to moderate aO2 tality rate among subjects with P in the 60–300 mm Hg stroke severity who were given oxygen.30 However, this aO2 range. The adjusted odds ratio for death was 1.2, although was not confirmed in a much larger study of stroke sub- the investigators concluded that “hyperoxia did not have a jects managed outside of critical care units (N ϭ 8,003 robust or consistently reproducible association with mortal- subjects).31 This study found that routine low-dose oxygen ity.” Curiously, for a study of the impact of hyperoxemia, treatment at 2–3 L/min from nasal cannula did not reduce they did not look at the average or highest oxygen level of (or increase) mortality or improve recovery even for sub- these subjects, which might have been more appropriate. El- jects with oxygen saturation in the range of 90–94%. mer and colleagues27 reported a small single-center study in 2015 in which severe hyperoxemia (P Ͼ 300 mm Hg or Traumatic Brain Injury Patients on Critical Care Units aO2 40 kPa) was associated with increased hospital mortality, 32 whereas moderate hyperoxemia (101–299 mm Hg) was not Rincon et al reported an increased risk of hospital death among 1,212 ventilated subjects with traumatic brain associated with decreased survival. A number of smaller stud- injury on critical care units if they were exposed to hy- ies have also been published, and a systematic review and peroxia (crude odds ratio (and adjusted odds ratio) for meta-analysis by Wang and colleagues20 concluded that hy- death was 1.5 if exposed to P Ͼ 300 mm Hg). Mortality peroxemia appears to be correlated with increased in-hospital aO2 was also increased with exposure to hypoxemia (crude mortality in survivors of adult cardiac arrest, although the odds ratio for hospital death was 2.3 for subjects with P results should be interpreted cautiously because of the sig- aO2 Ͻ60 mm Hg (7.99 kPa, equivalent to saturation of approx- nificant heterogeneity and the limited number of studies that imately 90–91%) or a P /F ratio Յ 300. were analyzed. aO2 IO2 Myocardial Infarction Ventilated General Critical Care Patients Patients with suspected myocardial infarction have been given supplemental oxygen on a routine basis for several 10 de Jonge et al reviewed the management of 36,307 ven- decades in the hope that this might increase oxygen de- tilated critical care subjects and reported increased mor- livery to the ischemic area of myocardium. As recently as tality (odds ratio ϭ 1.23) if the mean P in the first 24 h aO2 2010, an editorial in the British Medical Journal stated that Ͼ in hospital was 16.4 kPa (123 mm Hg). Optimal survival “it is reasonable to continue giving oxygen to people with was in the quintile of subjects with P ϭ 8.9–10.6 kPa 33 aO2 acute myocardial infarction.” This advice was given on (67–79 mm Hg), equivalent to saturation of approximately the basis that the increased mortality that was associated 93–96% at normal pH. Eastwood et al11 looked at the lowest with hyperoxia in a linked systematic review failed to oxygen level in the first 24 h (worst A-a gradient) for reach statistical significance. This suggestion (that a drug 152,680 ventilated critical care subjects with mean should be given to all patients with a common condition P ϭ 20 kPa (150 mm Hg). They reported an association of aO2 where there is no evidence of benefit but weak evidence of hypoxemia (but not hyperoxemia) with mortality. harm) must be unique to oxygen. Since that time, there A 2017 observational study by Helmerhorst et al28 re- have been 3 randomized trials of oxygen use in myocardial ported that mortality in ICU subjects was increased with infarction.34-36 None of these studies suggested any benefit marked hyperoxia (Ͼ 200 mm Hg or 27 kPa), and time for normoxic subjects, but the study by Stub and col- spent in hyperoxemia showed a linear relationship with leagues,35 which was undertaken in the pre-hospital envi- hospital mortality. Two recent systematic reviews with ronment with subjects with confirmed myocardial infarc- meta-analysis have addressed this subject.21,22 Both stud- tion, reported higher levels of cardiac enzymes and larger ies concluded that hyperoxemia was associated with in- infarct size on follow-up magnetic resonance scans in sub- creased hospital mortality, but caution is required in inter- jects randomized to receive supplemental oxygen. A 2018 preting the results because of heterogeneity in the included meta-analysis of these trials and all preceding trials found studies, including different definitions of hyperoxemia. no evidence of benefit from supplemental oxygen in nor-

RESPIRATORY CARE • OCTOBER 2019 VOL 64 NO 10 1297 OXYGEN USE IN CRITICAL ILLNESS moxic subjects with myocardial infarction but could not These pilot studies support the view that conservative rule out a harmful effect.37 It is also important to note that oxygen management may reduce mortality and duration of a proportion of subjects in some of these trials who were ventilation. Importantly, none of the 4 pilot studies raised randomized to air had already received oxygen in pre- any safety concerns related to the conservative use of ox- hospital care prior to randomization, so the true risks of ygen on critical care units. routine oxygen use in patients with suspected myocardial infarction may be underestimated in some of the included Randomized Trials of Oxygen Therapy in Critical studies. Care Settings

Pilot Studies of Conservative Oxygen Management A randomized trial of conservative oxygen therapy in for Ventilated Critical Care Patients critical care subjects with expected critical care unit stay of Ͼ 72 h was published in October 2016.43 This Italian There are 4 published pilot studies of conservative ox- study reported a critical care unit mortality rate of 11.6% ygen management in critical care subjects.38-41 All of these in subjects randomized to a target saturation range of 94– studies have demonstrated that such an approach is safe 98% (median P ϭ 11.6 kPa) compared with 20.2% and feasible. Suzuki and colleagues38 established in a be- aO2 critical care unit mortality with conventional therapy (me- fore-and-after pilot study of 105 subjects in 2014 that con- dian P ϭ 13.6 kPa). The absolute risk reduction was servative oxygen therapy for mechanically ventilated sub- aO2 8.6% (P ϭ .01) and the hospital mortality rate was 24.2% jects (target S ϭ 90–92%) was feasible and free of pO2 versus 33.9% (absolute risk reduction 9.9% (P ϭ .03). The adverse biochemical, physiological, or clinical outcomes authors urged caution in the interpretation of these results while allowing a marked decrease in excess oxygen expo- because the trial was terminated early due to slow recruit- sure. Within this study, the authors reported a lower inci- ment, although statistical significance was achieved for the dence of atelectasis on chest radiograph in subjects allo- primary and secondary end points. A larger trial of tar- cated to conservative oxygen therapy (odds ratio ϭ 0.28, geted oxygen use in critical care (ICU-ROX study) is due 95% CI 0.12–0.66, P ϭ .003) and earlier successful wean- to report in the very near future; the pilot study suggested ing from mechanical ventilator (adjusted ra- that critical care stay may be shortened by the use of tio ϭ 2.96, 95% CI 1.73–5.05, P Ͻ .001).38,39 Eastwood conservative oxygen therapy.44 and colleagues reported in a linked study of 100 subjects A randomized trial of hyperoxia in subjects with septic that a target oxygen saturation range of 88–92% was fea- shock showed a nonsignificant increase in 28-d mortality sible for mechanically ventilated cardiac arrest survivors associated with hyperoxia (43% vs 35%, P ϭ .12).45 Un- and resulted in more use of spontaneous ventilation mode fortunately, the trial was terminated prematurely because (ie, less mandatory ventilation), more ventilation on air of an increase in 2 secondary end points in the hyperoxia (F ϭ 0.21), and shorter stay on the critical care unit.40 IO2 group (ie, atelectasis and ICU weakness). Panwar et al41 completed a multi-center, randomized, controlled pilot trial of conservative (target range 88–92%) versus liberal (target range Ն 96%) oxygen therapy for Why Have Clinicians Used Too Much Oxygen for 103 mechanically ventilated subjects. They reported clear Many Decades? separation of oxygen levels, and although 90-d mortality was reduced in the conservative oxygen group (adjusted First, old habits die hard. Aggressive oxygen use for hazard ratio ϭ 0.77, 95% CI 0.40–1.5, P ϭ .44), this was seriously ill patients with unknown blood oxygen levels not significant in this small study. They did not identify was justified 100 years ago, but this is no longer true. any difference in the duration of mechanical ventilation or Tuberculosis and pneumonia were common in the early ventilator-free days between the subject groups. 20th century, and there were no effective treatments for Helmerhorst42 reported a before-and-after study in which either of these conditions. In these circumstances, correc- a target range was set at S ϭ 92–95% or P ϭ 55– tion of hypoxemia could, in some cases, prevent death pO2 aO2 86 mm Hg (7.3–11.5 kPa). They reported shorter periods until such time as the body’s natural defenses had over- of mechanical ventilation (an increase of 0.55 ventilator- come the pulmonary injuries. This also applied to victims free days) in phase 1 of implementation (95% CI 0.25– of gas attacks in World War I. At that time, there was no 0.84) and 0.44 d in phase 2 (95% CI 0.11–0.86). They also way to monitor blood oxygen levels, so clinicians admin- reported no difference in adjusted critical care unit mor- istered oxygen blindly. Precautionary oxygen use became tality or ventilator-free days, but hospital mortality de- standard practice, and this habit persisted long after the creased in reference to baseline. The adjusted odds ratio in introduction of routine blood gas sampling in the 1970s phases 1 and 2 were 0.84 (0.74–0.96) and 0.82 (0.69– and the availability of cheap and reliable oximeter probes 0.96), respectively. in the early 21st century.

1298 RESPIRATORY CARE • OCTOBER 2019 VOL 64 NO 10 OXYGEN USE IN CRITICAL ILLNESS

Another reason for this longstanding habit is that clini- high-concentration oxygen. This could be compared to cians are quick to take up new treatments, even if the costs keeping a gasoline pump running after the automatic cut- are high and the benefits are modest. This is especially true off point when filling your vehicle with gas. You might of products that are heavily promoted by pharmaceutical increase the amount of available gasoline marginally by companies, device manufacturers, and enthusiastic “pio- filling the inlet pipe as well as the tank, but you might also neering” clinicians. However, the same clinicians tend to spill a lot of gasoline on your shoes and possibly even be slow to dispense with traditional practices, even if they cause an explosion. have been proven to be ineffective or harmful. Unfortu- In addition to these reasons, oxygen was used for many nately, there are many levers and incentives to promote the decades for precautionary reasons in situations such as uptakes of new drugs and treatments, but there are few postoperative care, where some patients might suffer from levers to curtail outmoded practices. unexpected hypoxemia, sometimes with sudden onset. This Furthermore, simple physiology tends to be forgotten or practice may have been justified in the early 20th century ignored. Oxygen is a treatment for hypoxemia, not for when the blood oxygen level was unknown. Now that breathlessness. However, most lay people and many cli- oximeters are almost universally available, the precaution- nicians think that the symptom of breathlessness is caused ary use of oxygen may actually delay the recognition of mainly by low blood oxygen levels and that oxygen will clinical deterioration and may limit the treatment options. relieve this, even if the blood oxygen saturation is nor- As discussed by Downs3 in 2002, a patient who develops 46,47 mal. Disproving this fallacy requires only a volunteer respiratory deterioration while breathing air is likely to and a pulse oximeter. Very few humans can hold their have a gradual fall in routine S measurements, which pO2 breath long enough to produce a significant fall in their will alert the clinical team to the deterioration, and they blood oxygen saturation because our sense of breathless- may commence supplemental oxygen therapy to stabilize ness during breath-holding is caused mainly by failure to the patient while definitive treatment for underlying prob- expel our exhaust gas (carbon dioxide). A rising level of lems, such as pneumonia or heart failure, is commenced. carbon dioxide during breath-holding causes breathless- By contrast, if the same patient is using supplemental ox- ness long before the blood oxygen level falls. Ignorance of ygen on a precautionary basis, the S levels are likely to this fact has caused many deaths among free divers who pO2 remain reassuringly high even as the patient’s physiology may hyperventilate before a dive to prolong their dive time deteriorates. When the S does eventually fall, the pa- by taking in more oxygen. However, hyperventilation leads pO2 tient will be at a more advanced stage of deterioration, and to only about 2% rise in blood oxygen content due to a rise the therapeutic options are more limited because the pa- in oxygen saturation from 98% to 100% in a tient is now hypoxemic despite oxygen therapy and may typical healthy subject. This is in striking contrast to the require rapid transfer to the critical care unit. The 20th 47% reduction in blood carbon dioxide levels that can be achieved after just two minutes of voluntary hyperventi- century precaution has thus become a hazard to the 21st lation.48 This delays the onset of hypercapnia during the century patient. dive and allows a diver to stay underwater for longer be- Yet another reason that oxygen is overused is that ox- fore the onset of breathlessness. Unfortunately, a diver ygen therapy became normalized and institutionalized. Pa- who has hyperventilated may remain submerged long tients and many clinicians came to regard oxygen as part enough for hypoxia to develop. The first symptom of hyp- of standard medical care for every seriously ill patient, a oxia is disordered mental functioning (starting from satu- fashion that was likely influenced by the prominence of ration Ͻϳ80%), followed by the onset of unconscious- oxygen masks and tubes in medical television dramas and ness as the oxygen saturation falls to Ͻϳ56% in healthy in 20th-century movies. Even in the 21st century, there individuals.49-52 If this happens before surfaces, remains a general impression among patients and health the consequences are likely to be fatal. care professionals that oxygen is beneficial in most ill- Raising the arterial oxygen saturation above the physi- nesses.46,47 Breathless patients and almost all seriously ill ological range (about 96–98% for young adults and 94– patients have been conditioned to expect oxygen treatment 98% for older adults) has very little effect on blood oxy- and clinicians have tended to meet this expectation even gen content because, by definition, the hemoglobin when the patient’s oxygen saturation is already normal molecules in the blood are almost fully saturated with breathing air. oxygen at these saturation levels, and very little oxygen is The widespread belief that oxygen is always beneficial transported to the tissues by any other means.1 Giving is paralleled by widespread ignorance about the potential supplemental oxygen therapy to raise the P to side effects of oxygen therapy, until very recently.46,47 In aO2 Ͼ 100 mm Hg in a patient with S ϭ 98% can increase addition, clinicians like to be seen to do something, and it pO2 the oxygen saturation by only 2%, even if the P is has been said that clinicians are sometimes treating them- aO2 increased as high as 400 mm Hg by the administration of selves when they deliver a highly visible treatment such as

RESPIRATORY CARE • OCTOBER 2019 VOL 64 NO 10 1299 OXYGEN USE IN CRITICAL ILLNESS

Table 1. Key Principles of Oxygen Therapy Table 2. Suggested Oxygen Target Ranges for Patients With Serious Illness Who Are Not at Risk of Hypercapnia Oxygen is a treatment for hypoxemia. Giving oxygen does not relieve breathlessness, nor does it increase the oxygen supply to vital Suggested Target Range for Patients Who Are organs if the patient’s oxygen level is normal to start with. Guideline Aim for a normal or near-normal oxygen saturation level for most Not at Risk of Hypercapnia patients (94–98% or 92–96%). Aim for a lower level for those at risk from hypercapnia (88–92% or British Thoracic Society1 94–98% patient-specific range). Thoracic Society of Australia 92–96% Prescribers prescribe a ‘target range,’ and clinicians adjust equipment and New Zealand53 7 and flows to keep SpO2 within the target range. Siemieniuk et al 90–94% for most patients 90–92% for those with stroke or myocardial infarction oxygen to a critically ill patient irrespective of the oxim- etry reading, especially if breathlessness is present.47 Finally, from a medico-legal perspective, deaths due to lower with an upper limit of 96% for all patients on supple- hypoxia in situations such as disconnection of anesthetic mental oxygen therapy, a target range of 90–94% for most gas tubing or empty oxygen cylinders tend to be easy to patients, and a target range as low as 90–92% for some identify and often lead to medico-legal claims. However, conditions such as stroke or myocardial infarction. Some crit- deaths due to hyperoxemia are like deaths due to air pol- ical care specialists suggest aiming even lower for certain lution. There is extensive evidence that both of these fac- patientsrequiringmechanicalventilation,especiallythosewith tors can increase death rates in vulnerable populations. acute lung injury, but randomized clinical trials of this strat- Very few individual deaths are attributed to either hyper- egy are urgently required.12 oxia or air pollution, although it is likely that both of these factors cause the deaths of many thousands of Americans Immediate Management of Critical Illness in Pre- every year. If Chu and colleagues23 are correct, patients with Hospital Settings critical illness who have their blood oxygen saturation raised above 96% with supplemental oxygen may be 21% more The immediate management of acute illness requires likely to die than those treated with conservative oxygen rapid assessment of the patient and immediate life-saving therapy. Many clinicians continue to make their patients hy- measures such as airway management if appropriate. The peroxemic by administering oxygen on a precautionary basis, British Thoracic Society emergency oxygen guideline rec- possibly influenced in part by fear of censure or medico-legal ommends the use of high-concentration oxygen from a claims. It is rare for such clinicians to suffer any censure. By reservoir mask (or bag-valve-mask) during the process of contrast, most clinicians would be disciplined or worse if a assessment and resuscitation for critically ill patients. How- patient’s death was potentially the result of untreated hypox- ever, once the patient is found to have stable cardiac out- emia. Clinicians would also be criticized if they withheld put and a reliable oximetry trace can be obtained, clini- from critically ill patients a new treatment that reduced mor- cians should aim at a normal or near-normal oxygen tality by 21% but they are not presently criticized for admin- saturation range such as 94–98% or 92–96%, pending the istering an unnecessary treatment that may increase mortality availability of blood gas results.1,53 This advice also ap- by this amount. plies to survivors of cardiac arrest. The British Resuscita- tion Guideline recommends aiming for a saturation range Current Best Practice for Oxygen Use in Critical Illness of 94–98% once spontaneous cardiac output has returned because post-arrest hyperoxia has been associated with Overview increased mortality.54 There are some situations in which pulse oximetry is The consensus from most recent randomized trials, sys- unreliable, such as the patient being in a state of shock or tematic reviews, and meta-analyses is that supplemental ox- if the patient is thought to have carbon monoxide poison- ygen is indicated for patients with hypoxemia, but there is no ing. In these situations, high-concentration oxygen should evidence to support the use of oxygen for normoxemic pa- be administered until blood gas results (or carbon monox- tients (Table 1). The 2017 British Thoracic Society emer- ide measurements) are available in the hospital setting. In gency oxygen guideline advises aiming for a target saturation some pre-hospital situations, such as mountain rescue or of 94–98% for most patients with medical and surgical emer- cave rescue, oximeters may be unavailable or impractical, gencies.1 The Thoracic Society of Australia and New Zea- and high-concentration oxygen should be used for criti- land recommends a target range of 92–96% (Table 2).53 The cally ill patients in these situations until a reliable assess- more recent advice from Siemieniuk et al7 is to aim even ment of oxygenation can be made.

1300 RESPIRATORY CARE • OCTOBER 2019 VOL 64 NO 10 OXYGEN USE IN CRITICAL ILLNESS

Table 3. Oxygen Therapy Is Only One Element of Resuscitating a patients with acute stroke or acute myocardial infarction un- Critically Ill Patient less the saturation falls to Ͻ 92%, with some evidence that the best threshold at which to start oxygen therapy in these The oxygen carrying power of blood may be increased by: groups may be 90% or lower.7,23,53 The patient should be ⅐ Safeguarding the airway ⅐ Enhancing circulating volume continuously assessed using the ABCDE approach until sta- ⅐ Correcting severe anemia bilized, and a decision should be made as to whether the ⅐ Enhancing cardiac output patient can be appropriately managed in the current location ⅐ Avoiding/reversing respiratory depressants or if a higher level of care is required. ⅐ Giving supplemental oxygen if hypoxemic ⅐ Establishing the reason for hypoxia and treating the underlying cause (eg, heart failure or pneumonia) Oxygen Use in Critical Care Units

The vast majority of patients within the critical care unit Immediate Management of Critical Illness in will require supplementary oxygen to avoid hypoxemia, Hospital Settings but there is increasing evidence that clinical outcomes may be worsened with hyperoxemia. Again, in the critical care The immediate management of a patient who deteriorates unit it is recommended that, for patients receiving supple- suddenly in a hospital environment is similar to the immedi- mentary oxygen, an upper S limit of 96% should be set pO2 ate management of the pre-hospital patient discussed in the in acutely ill medical patients. In patients with stroke or previous section, but the key difference is that expert person- myocardial infarction, supplementary oxygen should not nel and blood gas measurements will be available much more be given if the S is Ͼ 92%.7 For patients at risk of pO2 quickly to guide optimal oxygen therapy (Table 3). In many hypercapnic respiratory failure, target S should be set at pO2 hospitals, the critical care outreach team will be called to 88–92%. These targets are likely to be adjusted in the assess patients with sudden hypoxemia or other critical ill- future based on the results of further studies. ness in the hospital. Supplementary oxygen is only required Oxygen-delivery systems in the critical care unit range if the S is Ͻ 94% in most patients, or if the S falls to from the traditional “standard-flow” systems, such as nasal aO2 aO2 Ͻ 88% in patients who are at risk of hypercapnic respiratory cannula, simple face mask, and reservoir mask, to high- failure.1 There is emerging evidence that an upper saturation flow humidified systems and oxygen delivered via nonin- limit of 96% should be applied to all acutely ill medical vasive and invasive ventilators. Standard-flow systems (ex- patients and that supplementary oxygen is not required in amples are shown in Fig. 1) deliver a flow 1–15 L/min,

Fig. 1. Devices that are commonly used to deliver oxygen. A: Non-rebreathing reservoir mask; B: simple face mask; C: Venturi mask; D: low-flow nasal cannula; E: low-flow humidified system (unheated); F: tracheostomy mask.

RESPIRATORY CARE • OCTOBER 2019 VOL 64 NO 10 1301 OXYGEN USE IN CRITICAL ILLNESS

is reduced initially until it reaches 0.4, at which point flow can then be reduced and a switch to conventional oxygen therapy considered.56 There are, however, contradictory findings around the use and efficacy of nasal high-flow oxygen, with inconclusive evidence of its superiority to conventional oxygen therapy and noninvasive ventilation in the reduction of intubation rates, mortality rates, and re-intubation rates.58-62 These conflicting findings may be attributed to the different HFNC application strategies and patient conditions investigated in the studies. Although it may offer benefit in terms of patient comfort, care must be taken to avoid any inappropriate delay in initiation of in- vasive ventilation caused by inappropriate perseverance with HFNC. Noninvasive and invasive mechanical ventilation will not be discussed in detail in this review; however, they allow the precise control of F , manipulation of ventila- IO2 tor settings allowing optimization of ventilation to achieve adequate arterial oxygenation. Regardless of the oxygen delivery system, the patient’s oxygen saturations are usually monitored continuously in critical care units and recorded hourly as part of routine observation, in addition to the ready availability of arterial blood samples from arterial lines and arterialized capillary Fig. 2. High-flow humidified system. A: Flow generator and hu- blood sampling. Ebmeier et al63 compared S and S in midification system; B: high-flow nasal cannula; C: high-flow nasal pO2 aO2 cannula on a patient. adult ICU patients, with the absence of a statistically sig- nificant bias in paired S /S supporting the use of oxi- pO2 aO2 metry to regulate oxygen therapy, although care was ad- which is insufficient to meet the patient’s flow require- vised when S recordings were 4.4% higher or lower aO2 ment, therefore causing room air to be entrained, thus than the S . Regular monitoring and frequent review by pO2 diluting the oxygen and reducing the amount of oxygen the multidisciplinary team should lead to precise titration delivered to the patient at alveolar level. The amount of of oxygen to maintain prescribed targets. Although targets entrained air is variable, dependent of the mask fit, pa- are regularly prescribed in the critical care unit, regular tient’s breathing frequency, , and tidal titration of oxygen does not always occur. This was clearly volume; therefore the amount of oxygen delivered can demonstrated in a study looking at goal-directed oxygen only be estimated. delivery based on S ; the investigators reported that the pO2 Oxygen therapy via high-flow nasal cannula (HFNC) S was in the prescribed range ϳ64% of the time.64 This pO2 has increased in popularity, and its use is becoming wide- was consistent with findings from the 2015 British Tho- spread in the critical care unit (Fig. 2). Oxygen and air is racic Society national audit of oxygen therapy in hospitals mixed in a blender, warmed to 37°C, and humidified, and in the United Kingdom (Table 4).65 then it is delivered to the patient at flows of 30–60 L/min. New systems for closed-loop control of oxygen ther- This is usually higher than the inspiratory flow of the apy have been developed; initial findings indicate an patient which reduces the likelihood of air entrainment and improvement in controlling target saturation compared allows the F to match that set at the blender. Several with manual titration; the use of these systems in clin- IO2 other effects have been documented, including pharyngeal ical practice is rare, however, and there is a need to washout, reduction of nasopharyngeal resis- investigate the safety and efficacy of these devices in tance, generation of positive expiratory pressure (3.2– future trials.66,67

7.4 cm H2O), and alveolar recruitment; furthermore, hu- midification may play a role in improved tolerance and Choice of Device improved mucociliary clearance by helping patients main- tain their secondary airway defense system.55–57 Because flow and F are independent, the F can be reduced The choice of device will depend on the patient’s med- IO2 IO2 separately from the flow. Typically the flow is set at 50 ical condition. Some common devices are shown in Figure L/min and the F is titrated to maintain target S .F 1 and Figure 2. Readers are referred to the British Thoracic IO2 pO2 IO2

1302 RESPIRATORY CARE • OCTOBER 2019 VOL 64 NO 10 OXYGEN USE IN CRITICAL ILLNESS

Table 4. UK Hospital Patients With an Oxygen Prescription* protocol development, interventions to improve ICU team- work, computer decision support, and behavioral economic 2008 2011 2012 2013 2015 interventions have been shown to assist in the implemen- Percent of UK hospital 17 14 14 14 14 tation of low tidal-volume ventilation.69 In addition, the patients using use of daily multidisciplinary team ward rounds with check- oxygen at time of lists, goal setting, and prompting have been shown to im- national audit prove compliance with set targets.72 It is likely that these Percent of patients 32 48 52 55 57 using oxygen who change principles can be utilized by respiratory therapists had a prescription to effect a change in the practice and culture around the Was oxygen signed for 520202128 delivery of oxygen therapy. on medicine rounds? Special Uses of Oxygen * The first national audit in 2008 was undertaken before the national oxygen guideline was published. The audit was repeated at intervals up to 2015. Oxygen as a Driving Gas for Nebulizers

Most patients with obstructive lung diseases such as Society oxygen guideline for advice regarding when to use asthma, COPD, or bronchiectasis have additional air- each type of mask or device and for advice regarding when flow obstruction during exacerbations and require bron- to use humidified oxygen.1 chodilator treatment. Nebulized bronchodilators such as salbutamol are widely used to treat this group of pa- Role of the Respiratory Therapist tients. For patients with asthma, it is recommended that the driving gas should be oxygen; however, many pa- As evidence-based health care becomes more available, tients with COPD and some patients with bronchiectasis one would assume that this would translate into enhanced are at risk of hypercapnia, and the recommendation for patient care and improved outcomes; nevertheless, patients such patients is to use an air-driven nebulizer with sup- fail to receive care based on the best available evidence or plemental oxygen via nasal cannula if necessary to main- current guidelines.68 Within health care, and also within tain a target saturation in the range of 88–92%.1 If an critical care, it is well recognized that the process of trans- air-driven or ultrasonic nebulizer is not available, cli- lating evidence into practice is slow, and it is difficult to nicians can use multiple doses of bronchodilator using a create a culture maintaining a change once it has been metered-dose inhaler and a spacer. If an oxygen-driven implemented. This has been well demonstrated within the nebulizer must be used, the advice is to limit this to field of critical care in the implementation of strategies 6 min, which will deliver most of the bronchodilator such as low tidal-volume ventilation in the management of drug, although there will still be some rise in the blood ARDS, which, similarly to oxygen titration, requires a carbon dioxide level.1,73 change in mindset rather than additional resources or equip- ment.69 Carbon Monoxide Poisoning A significant barrier to change in culture is a lack of knowledge of the current literature and a lack of awareness The hemoglobin molecule has a much greater affinity of guidelines, often with staff believing they are meeting for carbon monoxide than for oxygen (and it displaces guidelines despite clinical practice differing from recom- oxygen). Unfortunately, most pulse oximeters miscat- mendations.70 Kelly and Maddon,70 in their study looking egorize as oxyhemoglobin, and the at how health care professionals perceive oxygen therapy, oximeter reading may be misleadingly normal despite reported inconsistency in reported beliefs, understanding, significant arterial hypoxemia. This problem is further and clinical practice. compounded by a normal blood oxygen tension in these Respiratory therapists, as specialists in respiratory care, circumstances. It is vitally important to measure the possess increased knowledge and understanding around blood carbon monoxide level in all patients who might the prescription, delivery, and titration of oxygen therapy. possibly have inhaled this gas and to give high-concen- Respiratory therapists have been shown to improve com- tration oxygen to accelerate the clearance of carbon pliance in the delivery of protocol-driven care and in the monoxide from the hemoglobin molecule. This is one of implementation of standard guidelines. In the field of me- the rare situations in which the pulse oximetry target is chanical ventilation, RTs are instrumental in the delivery 100% until the blood carbon monoxide level has fallen of ventilation and weaning.71 As such, RTs are in an ideal to safe levels. position to implement reform in targeted oxygen delivery. Hyperbaric oxygen may also be beneficial in severe Strategies such as audit and feedback, provider education, cases of carbon monoxide poisoning. Although hyperbaric

RESPIRATORY CARE • OCTOBER 2019 VOL 64 NO 10 1303 OXYGEN USE IN CRITICAL ILLNESS oxygen accelerates the clearance of carbon monoxide from is important not to swing too far in the opposite direction the bloodstream, the evidence for improved clinical out- and thus ignore the dangers of hypoxemia. National audits comes is weak but increasing.74,75 Hyperbaric oxygen is in the United Kingdom have shown a very slow improve- outside the scope of this review, except to say that it is ment in the proportion of patients using oxygen who have considered to be effective in the management of decom- a prescription since national guidelines were first pub- pression illness or arterial gas embolism, and it is some- lished in 2008 (Table 4).65 times used to accelerate recovery from carbon monoxide It is already becoming evident that early 21st-century poisoning. Most other uses are regarded as experimental at advice to aim for normal or near-normal oxygen saturation the present time. ranges such as 94–98% for most seriously ill patients was probably overly generous.1 It may be harmful to push the Medically Managed Pneumothorax S above 96% with the use of supplemental oxygen, and pO2 lower target ranges such as 92–96% or even 90–94% have Pneumothorax is usually managed either by aspiration been proposed.23,53 It is likely that future research will or by intercostal tube drainage. However, a small primary define optimal target ranges for specific medical condi- spontaneous pneumothorax may be allowed to self-absorb. tions, just as there is consensus at present that most pa- In cases where a patient wishes to avoid drainage or if tients with a history of hypercapnia should have a target there is a contra-indication to drainage, the application of range of 88–92%.1 high-concentration oxygen will expedite the clearance of The situation in critical care units is even more complex air from the pleural cavity.1 than in the emergency room or the admissions unit. The present consensus is to aim for saturations in the low 90s Oxygen and Avoidance of Perioperative Wound for most ventilated patients, and some authors propose Infection permissive hyperoxemia at lower levels, especially for pa- tients with acute lung injury; clinical trials of this strategy High of oxygen have been given to post- are required as a matter of some urgency.6,13 It is likely operative patients in the hope that this may reduce the risk that future studies will define optimal target ranges (and of wound infection. Meta-analyses of this topic have largely optimal ventilation strategies) for different subgroups of been negative, with a suggestion that there may be some ventilated patients on critical care units. In the meantime, benefit in some subgroups.76 The largest such study actu- there is a great deal of evidence that untreated hypoxemia ally suggested a delayed mortality risk in the group who and iatrogenic hyperoxemia are both harmful and should received high-concentration oxygen.77,78 both be avoided. The former hazard is much feared and very rare, but the latter hazard is still poorly recognized Future Directions and very common.

Although oxygen is an important treatment for hypox- emia, most available evidence suggests that oxygen has REFERENCES been overused for many decades and is still widely over- 1. O’Driscoll BR, Howard LS, Earis J, Mak V; British Thoracic Society used. There is no evidence of benefit from hyperoxemia in Emergency Oxygen Guideline Group; BTS Emergency Oxygen most medical conditions, but there is increasing evidence Guideline Development Group. BTS guideline for oxygen use in of harm from excessive oxygen use. If the calculations of adults in healthcare and emergency settings. Thorax 2017;72(Suppl Chu and colleagues23 are correct, it is possible that many 1):ii1-ii90. thousands of deaths in the United States every year are 2. Lanman JT, Guy LP, Dancis J. Retrolental fibroplasia and oxygen therapy. J Am Med Assoc 1954;155(3):223-226. caused by excessive oxygen therapy but not recognized as 3. Downs JB. Has oxygen administration delayed appropriate respira- such. tory care? Fallacies regarding oxygen therapy. Respir Care 2003; The biggest challenge regarding oxygen use at present 48(6):611-620. is to change the hearts and minds of the public and of 4. Hale KE, Gavin C, O’Driscoll BR. Audit of oxygen use in emer- health care professionals so that they recognize oxygen as gency ambulances and in a hospital emergency department. Emerg Med J 2008;25(11):773-776. a useful drug for a specific indication (hypoxemia), and 5. Smith GB, Prytherch DR, Watson D, Forde V, Windsor A, Schmidt not as a routine and essential measure to be applied with- PE, et al. SpO(2) values in acute medical admissions breathing air– out thought to all patients with serious medical conditions. implications for the British Thoracic Society guideline for emer- The first requirement for this change will be recognition gency oxygen use in adult patients? Resuscitation 2012;83(10):1201- that oxygen is a drug, not a panacea. Like all drugs, it 1205. 6. Slutsky AS. Consensus conference on mechanical ventilation–Jan should be prescribed within its therapeutic range, and it 28-30, 1993 at Northbrook, Illinois, USA. Part I. European Society can have serious side effects and can cause deaths if mis- of Intensive Care Medicine, the ACCP and the SCCM. Intensive used. However, like all “see-saw” problems in medicine, it Care Med 1994;20(1):64-79.

1304 RESPIRATORY CARE • OCTOBER 2019 VOL 64 NO 10 OXYGEN USE IN CRITICAL ILLNESS

7. Siemieniuk RAC, Chu DK, Ha-Yeon Kim L, Guell-Rous MR, Al- 25. Kilgannon JH, Jones AE, Parrillo JE, Dellinger RP, Milcarek B, hazanni W, et al. Oxygen therapy for acutely ill medical patients: a Hunter K, et al. Relationship between supranormal oxygen tension clinical practice guideline. BMJ 2018;363:k41069. and outcome after resuscitation from cardiac arrest. Circulation 2011; 8. Kelly FE, Fong K, Hirsch N, Nolan JP. Intensive care medicine is 60 123(23):2717-2722. years old: the history and future of the intensive care unit. Clin Med 26. Bellomo R, Bailey M, Eastwood GM, Nichol A, Pilcher D, Hart GK, (Lond) 2014;14(3):376-379. et al. Arterial hyperoxia and in-hospital mortality after resuscitation 9. Rachmale S, Li G, Wilson G, Malinchoc M, Gajic O. Practice of from cardiac arrest. Crit Care 2011;15(2):R90. excessive FIO2 and effect on pulmonary outcomes in mechanically 27. Elmer J, Scutella M, Pullalarevu R, Wang B, Vaghasia N, Trzeciak ventilated patients with acute lung injury. Respir Care 2012;57(11): S; Pittsburgh Post-Cardiac Arrest Service (PCAS). The association 1887-93. between hyperoxia and patient outcomes after cardiac arrest: analy- 10. de Jonge E, Peelen L, Keijzers PJ, Joore H, de Lange D, van der sis of a high-resolution database. Intensive Care Med 2015;41(1): Voort PH. Association between administered oxygen, arterial partial 49-57. oxygen pressure and mortality in mechanically ventilated intensive 28. Helmerhorst HJ, Arts DL, Schultz MJ, van der Voort PH, Abu- care unit patients. Crit Care 2008;12(6):R156. Hanna A, de Jonge E, van Westerloo DJ. Metrics of arterial hyper- 11. Eastwood G, Bellomo R, Bailey M, Taori G, Pilcher D, Young P, oxia and associated outcomes in critical care. Crit Care Med 2017; Beasley R. Arterial oxygen tension and mortality in mechanically 45(1):187-195. ventilated patients. Intensive Care Med 2012;38(1):91-98. 29. Rincon F, Kang J, Maltenfort M, Vibbert M, Urtecho J, Athar 12. O’Driscoll BR, Dark P, Wijesinghe M, McAuley D. Oxygen use is MK, et al. Association between hyperoxia and mortality after becoming more conservative on intensive care units in the UK. Tho- stroke: a multicenter cohort study. Crit Care Med 2014;42(2): rax 2016;71(Suppl 3):A176. 387-396. 13. Martin DS, Grocott MPW. Oxygen therapy in critical illness: precise 30. Rønning OM, Guldvog B. Should stroke victims routinely receive control of arterial oxygenation and permissive hypoxemia. Crit Care supplemental oxygen? A quasi-randomized controlled trial. Stroke Med 2013;41(2):423-432. 1999;30(10):2033-2037. 14. Smith JL. The pathological effects due to increase of oxygen tension 31. Roffe C, Nevatte T, Sim J, Bishop J, Ives N, Ferdinand P, et al. in the air breathed. J Physiol 1899;24(1):19-35. Effect of routine low-dose oxygen supplementation on death and 15. Freeman BA, Crapo JD. Hyperoxia increases oxygen radical pro- disability in adults with acute stroke: the Stroke Oxygen Study duction in rat lungs and lung mitochondria. J Biol Chem 1981; Randomized Clinical trial. JAMA 2017;318(12):1125-1135. 256(21):10986-10992. 32. Rincon F, Kang J, Vibbert M, Urtecho J, Athar MK, Jallo J. Signif- 16. Barazzone C, Horowitz S, Donati YR, Rodriguez I, Piguet PF. Ox- icance of arterial hyperoxia and relationship with case fatality in ygen toxicity in mouse lung: pathways to cell death. Am J Respir traumatic brain injury: a multicentre cohort study. J Neurol Neuro- Cell Mol Biol 1998;19(4):573-581. surg Psychiatry 2014;85(7):799-805. 17. Griffith DE, Holden WE, Morris JF, Min LK, Krishnamurthy. Ef- 33. Atar D. Should oxygen be given in myocardial infarction? BMJ fects of common therapeutic concentrations of oxygen on lung clear- 2010;340:c3287. ance of 99mTc DTPA and bronchoalveolar lavage albumin concen- 34. Ranchord AM, Argyle R, Beynon R, Perrin K, Sharma V, Weatherall tration. Am Rev Respir Dis 1986;134(2):233-237. M, et al. High-concentration versus titrated oxygen therapy in ST- 18. Farquhar H, Weatherall M, Wijesinghe M, Perrin K, Ranchord A, elevation myocardial infarction: a pilot randomized controlled trial. Simmonds M, Beasley R. Systematic review of studies of the effect Am Heart J 2012;163(2):168-175. of hyperoxia on coronary blood flow. Am Heart J 2009;158(3):371- 35. Stub D, Smith K, Bernard S, Nehme Z, Stephenson M, Bray JE, et 377. al.; AVOID Investigators. Air versus oxygen in ST-segment-eleva- 19. Thomson AJ, Webb DJ, Maxwell SR, Grant IS. Oxygen therapy in tion myocardial infarction. Circulation 2015;131(24):2143-2150. acute medical care. BMJ 2002;324(7351):1406-1047. 36. Hofmann R, James SK, Jernberg T, Lindahl B, Erlinge D, Witt N, et 20. Wang CH, Chang WT, Huang CH, Tsai MS, Yu PH, Wang AY, et al. Oxygen therapy in suspected acute myocardial infarction. N Engl al. The effect of hyperoxia on survival following adult cardiac arrest: J Med 2017;377(13):1240-1249. a systematic review and meta-analysis of observational studies. Re- 37. Sepehrvand N, James SK, Stub D, Khoshnood A, Ezekowitz JA, suscitation 2014;85(9):1142-1148. Hofmann R. Effects of supplemental oxygen therapy in patients with 21. Damiani E, Adrario E, Girardis M, Romano R, Pelaia P, Singer suspected acute myocardial infarction: a meta-analysis of randomised M, Donati A. Arterial hyperoxia and mortality in critically ill clinical trials. Heart 2018;104(20):1691-1698. patients: a systematic review and meta-analysis. Crit Care 2014; 38. Suzuki S, Eastwood GM, Glassford NJ, Peck L, Young H, Garcia- 18(6):711. Alvarez M, et al. Conservative oxygen therapy in mechanically ven- 22. Helmerhorst HJ, Roos-Blom MJ, van Westerloo DJ, de Jonge E. tilated patients: a pilot before-and-after trial. Crit Care Med 2014; Association between arterial hyperoxia and outcome in subsets of 42(6):1414-1422. critical illness: a systematic review, meta-analysis, and meta- 39. Suzuki S, Eastwood GM, Goodwin MD, Noe¨ GD, Smith PE, Glass- regression of cohort studies. Crit Care Med 2015;43(7):1508- ford N, et al. Atelectasis and mechanical ventilation mode during 1519. conservative oxygen therapy: a before-and-after study. J Crit Care 23. Chu DK, Kim LHY, Young PJ, Zamiri N, Almenawer SA, Jae- 2015;30(6):1232-1237. schke R, et al. Mortality and morbidity in acutely ill adults treated 40. Eastwood GM, Tanaka A, Espinoza ED, Peck L, Young H, Mårtens- with liberal versus conservative oxygen therapy (IOTA): A sys- son J, et al. Conservative oxygen therapy in mechanically ventilated tematic review and meta-analysis. Lancet 2018;391(10131):1693- patients following cardiac arrest: a retrospective nested cohort study. 1705. Resuscitation 2016;101(1):108-114. 24. Kilgannon JH, Jones AE, Shapiro NI, Angelos MG, Milcarek B, 41. Panwar R, Hardie M, Bellomo R, Barrot L, Eastwood GM, Young PJ Hunter K, et al. Association between arterial hyperoxia following et al. Conservative versus liberal oxygenation targets for mechani- resuscitation from cardiac arrest and in-hospital mortality. JAMA cally ventilated patients: a pilot multicenter randomized controlled 2010;303(21):2165-2171. trial. Am J Respir Crit Care Med 2016;193(1):43-51.

RESPIRATORY CARE • OCTOBER 2019 VOL 64 NO 10 1305 OXYGEN USE IN CRITICAL ILLNESS

42. Helmerhorst HJ, Schultz MJ, van der Voort PH, Bosman RJ, Juffer- tive pressure ventilation: a systematic review and meta-analysis. Am J mans NP, de Wilde RB, et al. Effectiveness and clinical outcomes of Emerg Med 2018;36(2):226-233. a two-step implementation of conservative oxygenation targets in 60. Ou X, Hua Y, Liu J, Gong C, Zhao W. Effect of high-flow nasal critically ill patients: a before and after trial. Crit Care Med 2016; cannula oxygen therapy in adults with acute hypoxemic respiratory 44(3):554-63. failure: a meta-analysis of randomized controlled trials. CMAJ 2017; 43. Girardis M, Busani S, Damiani E, Donati A, Rinaldi LM, Morelli A, 189(7):260-267. et al. Effect of conservative vs conventional oxygen therapy on 61. Hernandez G, Vaquero C, Colinas L, Cuena R, Gonzalez P, Canabal mortality amongst patients in an Intensive Care Unit. JAMA 2016; A, et al. Effect of postextubation high-flow nasal cannula vs nonin- 316(15):1583-1589. vasive ventilation on reintubation and postextubation respiratory fail- 44. Young PJ, Mackle DM, Bailey MJ, Beasley RW, Bennett VL, Deane ure in high-risk patients: a randomized clinical trial. JAMA 2016; AM, et al. Intensive care unit randomised trial comparing two ap- 316(15):1565-1574. proaches to oxygen therapy (ICU-ROX): results of the pilot phase. 62. Nedel WL, Deutschendorf C, Filho EMR. High-flow nasal cannula Crit Care Resusc 2017;19(4):344-354. in critically ill subjects with or at risk for respiratory failure: a 45. Asfar P, Schortgen F, Boisrame´-Helms J, Charpentier J, Guerot E, systematic review and meta-analysis. Respir Care 2017;62(1):123- Megarbane B, et al. Hyperoxia and hypertonic saline in patients 132. with septic shock (HYPERS2S): a two-by-two factorial, multi- 63. Ebmeier SJ, Barker M, Bacon M, Beasley RC, Bellomo R, Knee C, centre, randomised, clinical trial. Lancet Respir Med 2017;5(3): et al. A two-centre observational study of simultaneous pulse oxim- 180-190. etry and arterial oxygen saturation recordings in intensive care pa- 46. O’Driscoll BR, Bakerly ND, Caress AL, Roberts J, Gaston M, New- tients. Anaesth Inten Care 2018;46(3):297-303. ton M, Yorke J. A study of attitudes, beliefs and organisational 64. Jochmans S, Vong LV, Rolin N, Sy O, Chelly J, Ellrodt O, et al. barriers related to safe emergency oxygen therapy for patients with Efficiency of goal-directed oxygen delivery in ICU patients. Anaes- COPD (chronic obstructive pulmonary disease) in clinical practice thesiol Intensive Ther 2016;48(3):151-157. and research. BMJ Open Respir Res 2016;3:e000102. 65. O’Driscoll BR. British Thoracic Society Emergency Oxygen Audit 47. Kelly CA, Lynes D, O’Brien MR, Shaw B. A wolf in sheep’s cloth- Report 2015. London: British Thoracic Society, 2015. Available at: ing? Patients’ and healthcare professionals’ perceptions of oxygen https://www.brit-thoracic.org.uk/document-library/quality- therapy: an interpretative phenomenological analysis. Clin Respir J improvement/audit-reports/emergency-oxygen-2015. Accessed July 2018;12(2):616-632. 13, 2019. 48. Rotsztain A, Haddad R, Canter HG. Blood gas changes during vol- 66. Hansen EF, Hove JD, Bech CS, Jensen JS, Kallemose T, Vestbo J. untary hyperventilation in normal and disease states. Am Rev Respir Automated oxygen control with O2matic during admission with ex- Dis 1970;102(2):205-212. acerbation of COPD. Int J Chron Obstruct Pulmon Dis 2018;13: 49. Hoffman CE, Clark RT Jr, Brown EB Jr. Blood oxygen saturations 3997-4003. and duration of consciousness in anoxia at high altitudes. Am J 67. L’Her E, Jaber S, Verzilli D, Jacob C, Huiban B, Futier E, et al. Physiol 1946;145(3):685-692. Automated oxygen administration versus conventional oxygen ther- 50. Comroe JH Jr, Bahnson ER, Coates EO Jr. Mental changes occurring apy after major abdominal or thoracic surgery: study protocol for an in chronically anoxemic patients during oxygen therapy. J Am Med international multicentre randomised controlled study. BMJ Open Assoc 1950;143(12):1044-1048. 2019;9(1):e023833. 51. Boycott AE, Haldane JS. The effects of low atmospheric pressures 68. Pierson DJ. Translating evidence into practice. Respir Care 2009; on respiration. J Physiol 1908;37(5-6):355-377. 54(10):1386-1401. 52. Yoneda I, Tomoda M, Tokumaru O, Sato T, Wantanabe Y. Time of 69. Sjoding MW. Translating evidence into practice in ARDS: team- useful consciousness determination in aircrew members with refer- work, clinical decision support, and behavioural economic interven- ence to prior altitude chamber experience and age. Aviat Space tions. Curr Opin Crit Care 2017;23(5):406-411. Environ Med 2000;71(2):72-76. 70. Kelly CA, Madden M. How do health-care professionals perceive 53. Beasley R, Chien J, Douglas J, Eastlake L, Farah C, King G, et al. oxygen therapy? A critical interpretive synthesis of the literature. Thoracic Society of Australia and New Zealand oxygen guidelines Chron Respi Dis 2015;12(1):11-23. for acute oxygen use in adults: ‘swimming between the flags’. Re- 71. Karthika M, Wong D, Nair SG, Pillai LV, Matthew CS. Lung ul- spirology 2015;20(8):1182-1191. trasound: the emerging role of respiratory therapists. Respir Care 54. Soar J, Deakin C, Lockey A, Nolan J, Perkins G. Resuscitation 2019;64(2):217-229. Guidelines 2015 Adult Life Support. London: Resuscitation Council 72. Cavalcanti AB, Bozza FA, Machado FR, Salluh JI, Campagnucci UK, 2015. VP, Vendramim P, et al. Effect of a quality improvement interven- 55. Zhang J, Lin L, Pan K, Zhou J, Huang X. High-flow nasal cannula tion with daily round , goal setting, and clinician prompt- therapy for adult patients. J Int Med Res 2016;44(6):1200-1211. ing on mortality of critically ill patients: a randomized clinical trial. 56. Levy SD, Alladina JW, Hibbert KA, Harris RS, Bajwa EK, Hess DR. JAMA 2016;315(14):1480-1490. High-flow oxygen therapy and other inhaled therapies in intensive 73. Bardsley G, Pilcher J, McKinstry S, Shirtcliffe P, Berry J, Fingleton care units. Lancet 2016;387(10030):1867-1878. J, et al. Oxygen versus air-driven nebulisers for exacerbations of 57. Ashraf-Kashani N, Kumar R. High-flow nasal oxygen therapy. BJA chronic obstructive pulmonary disease: a randomised controlled trial. Educ 2017;17:57-62. BMC Pulm Med 2018;18(1):157. 58. Monro-Somerville T, Sim M, Ruddy J, Vilas M, Gillies MA. The 74. Buckley NA, Juurlink DN, Isbister G, Bennett MH, Lavonas EJ. Hy- effect of high-flow nasal cannula oxygen therapy on mortality and perbaric oxygen for carbon monoxide poisoning. Cochrane Database intubation rate in acute respiratory failure: a systematic review and Syst Rev 2011;4:CD002041. meta-analysis. Crit Care Med 2017;45(4):e449–e456. 75. Lin CH, Su WH, Chen YC, Feng PH, Shen WC, Ong JR, et al. 59. Ni YN, Luo J, Yu H, Liu D, Liang BM, Liang ZA. The effect of Treatment with normobaric or hyperbaric oxygen and its effect on high-flow nasal cannula in reducing the mortality and the rate of neuropsychometric dysfunction after carbon monoxide poisoning: a endotracheal intubation when used before mechanical ventilation systematic review and meta-analysis of randomized controlled trials. compared with conventional oxygen therapy and noninvasive posi- Medicine (Baltimore) 2018;97(39):e12456.

1306 RESPIRATORY CARE • OCTOBER 2019 VOL 64 NO 10 OXYGEN USE IN CRITICAL ILLNESS

76. de Jonge S, Egger M, Latif A, Loke YK, Berenholtz S, Boermeester surgery: the PROXI randomized clinical trial. JAMA 2009;302(14): M, et al. Effectiveness of 80% vs 30-35% fraction of inspired oxygen 1543-1550. in patients undergoing surgery: an updated systematic review and 78. Meyhoff CS, Jorgensen LN, Wetterslev J, Christensen KB, Rasmus- meta-analysis. Br J Anaesth 2019;122(3):325-334. sen LS; PROXI Trial Group. Increased long-term mortality after a 77. Meyhoff CS, Wetterslev J, Jorgensen LN, Henneberg SW, Høgdall high perioperative inspiratory oxygen fraction during abdominal sur- C, Lundvall L, et al. Effect of high perioperative oxygen fraction on gery: follow-up of a randomized clinical trial. Anesth Analg 2012; surgical site infection and pulmonary complications after abdominal 115(4):849-854.

RESPIRATORY CARE • OCTOBER 2019 VOL 64 NO 10 1307